-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathserver.py
188 lines (164 loc) · 7.55 KB
/
server.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
import argparse, os
import cv2
import torch
import numpy as np
from omegaconf import OmegaConf
from PIL import Image
from tqdm import tqdm, trange
from imwatermark import WatermarkEncoder
from itertools import islice
from einops import rearrange
import time
import onnxruntime
from pytorch_lightning import seed_everything
from torch import autocast
from contextlib import nullcontext
from ldm.util import instantiate_from_config
from ldm.models.diffusion.plms import PLMSSampler
from diffusers.pipelines.stable_diffusion.safety_checker import StableDiffusionSafetyChecker
from transformers import AutoFeatureExtractor
from flask import Flask, request, jsonify
app = Flask(__name__)
def load_model_from_config(config, ckpt, verbose=False):
if ckpt.endswith(".ckpt"):
print(f"Loading model from {ckpt}")
pl_sd = torch.load(ckpt, map_location="cpu")
if "global_step" in pl_sd:
print(f"Global Step: {pl_sd['global_step']}")
sd = pl_sd["state_dict"]
model = instantiate_from_config(config.model)
m, u = model.load_state_dict(sd, strict=False)
if len(m) > 0 and verbose:
print("missing keys:")
print(m)
if len(u) > 0 and verbose:
print("unexpected keys:")
print(u)
elif ckpt.endswith(".safetensors"):
print(f"Loading model from {ckpt}")
from safetensors.torch import load_file
sd = load_file(ckpt)
model = instantiate_from_config(config.model)
m, u = model.load_state_dict(sd, strict=False)
if len(m) > 0 and verbose:
print("missing keys:")
print(m)
if len(u) > 0 and verbose:
print("unexpected keys:")
print(u)
else:
raise ValueError(f"Unknown extension for checkpoint {ckpt}")
model.cuda()
model.eval()
return model
def put_watermark(img, wm_encoder=None):
if wm_encoder is not None:
img = cv2.cvtColor(np.array(img), cv2.COLOR_RGB2BGR)
img = wm_encoder.encode(img, 'dwtDct')
img = Image.fromarray(img[:, :, ::-1])
return img
def load_replacement(x):
try:
hwc = x.shape
y = Image.open("assets/rick.jpeg").convert("RGB").resize((hwc[1], hwc[0]))
y = (np.array(y)/255.0).astype(x.dtype)
assert y.shape == x.shape
return y
except Exception:
return x
def numpy_to_pil(images):
"""
Convert a numpy image or a batch of images to a PIL image.
"""
if images.ndim == 3:
images = images[None, ...]
images = (images * 255).round().astype("uint8")
pil_images = [Image.fromarray(image) for image in images]
return pil_images
def check_safety(x_image):
safety_checker_input = safety_feature_extractor(numpy_to_pil(x_image), return_tensors="pt")
x_checked_image, has_nsfw_concept = safety_checker(images=x_image, clip_input=safety_checker_input.pixel_values)
assert x_checked_image.shape[0] == len(has_nsfw_concept)
for i in range(len(has_nsfw_concept)):
if has_nsfw_concept[i]:
x_checked_image[i] = load_replacement(x_checked_image[i])
return x_checked_image, has_nsfw_concept
def load_stable_diffusion(config_path, ckpt_path):
# if not os.path.exists(ckpt_path):
# print("Downloading model checkpoint...")
# os.system(f"wget https://huggingface.co/runwayml/stable-diffusion-v1-5/resolve/main/v1-5-pruned-emaonly.ckpt?download=true -O {ckpt_path}")
config = OmegaConf.load(config_path)
model = load_model_from_config(config, ckpt_path)
device = torch.device("cuda") if torch.cuda.is_available() else torch.device("cpu")
sampler = PLMSSampler(model)
model.to(device)
return model, sampler
def parse_args():
parser = argparse.ArgumentParser("Stable Diffusion Inference")
parser.add_argument("--port", type=int, default=8000, help="Port to run the server on")
parser.add_argument("--ckpt", type=str, required = True, help="Path to the model checkpoint")
parser.add_argument("--verifier", type=str, required = True, help="Path to the verifier model")
return parser.parse_args()
args = parse_args()
safety_model_id = "CompVis/stable-diffusion-safety-checker"
safety_feature_extractor = AutoFeatureExtractor.from_pretrained(safety_model_id)
safety_checker = StableDiffusionSafetyChecker.from_pretrained(safety_model_id)
wm = "StableDiffusionV1"
wm_encoder = WatermarkEncoder()
wm_encoder.set_watermark('bytes', wm.encode('utf-8'))
verifier = onnxruntime.InferenceSession(args.verifier)
model, sampler = load_stable_diffusion("inference_config.yaml", args.ckpt)
@app.route('/')
def index():
return "Hello, World!"
@app.route('/generate_image', methods=['POST'])
def generate_image():
seed_everything(42)
json_request = request.get_json(force=True)
output_path = json_request['output_path']
prompt = json_request['prompt']
requested_C = json_request['C']
requested_H = json_request['H']
requested_W = json_request['W']
requested_f = json_request['f']
requested_ddim_steps = json_request['ddim_steps']
requested_scale = json_request['scale']
start = time.time()
start_code = None
precision_scope = autocast
with torch.no_grad():
with precision_scope("cuda"):
with model.ema_scope():
for prompts in tqdm([[prompt]], desc="data"):
uc = None
if requested_scale != 1.0:
uc = model.get_learned_conditioning(1 * [""])
if isinstance(prompts, tuple):
prompts = list(prompts)
c = model.get_learned_conditioning(prompts)
shape = [requested_C, requested_H // requested_f, requested_W // requested_f]
samples_ddim, _ = sampler.sample(S=requested_ddim_steps,
conditioning=c,
batch_size=1,
shape=shape,
verbose=False,
unconditional_guidance_scale=requested_scale,
unconditional_conditioning=uc,
eta=0.0,
x_T=start_code)
x_samples_ddim = model.decode_first_stage(samples_ddim)
x_samples_ddim = torch.clamp((x_samples_ddim + 1.0) / 2.0, min=0.0, max=1.0)
x_samples_ddim = x_samples_ddim.cpu().permute(0, 2, 3, 1).numpy()
x_checked_image, has_nsfw_concept = check_safety(x_samples_ddim)
x_checked_image_torch = torch.from_numpy(x_checked_image).permute(0, 3, 1, 2)
x_checked_image_numpy = x_checked_image_torch.cpu().numpy()
print(x_checked_image_numpy.shape)
verified_embedding = verifier.run(None, {'input': x_checked_image_numpy.astype(np.float32)})[0].tolist()
print(len(verified_embedding))
x_sample = 255. * rearrange(x_checked_image_numpy[0], 'c h w -> h w c')
img = Image.fromarray(x_sample.astype(np.uint8))
img = put_watermark(img, wm_encoder)
img.save(output_path)
return jsonify({"output_path": output_path, "embedding": verified_embedding, "time": time.time() - start})
if __name__ == "__main__":
app.run(host='127.0.0.1', port = args.port)